Spreads Enriched with Three Different Levels of Vegetable Oil Sterols And

Home , Carotenoid

Spreads enriched with three different levels of vegetable oil sterols and the degree of cholesterol lowering in normocholesterolaemic and mildly hypercholesterolaemic subjects

HFJ Hendriks1*, JA Weststrate2, T van Vliet1 and GW Meijer2

1TNO Nutrition and Food Research Institute, Zeist, The Netherlands; and 2The Unilever Nutrition Center, Unilever Research, Vlaardingen, The Netherlands

Objective: To investigate the dose-response relationship between cholesterol lowering and three different, relatively low intake levels of plant sterols (0.83, 1.61, 3.24 g=d) from spreads. To investigate the effects on lipid- soluble (pro)vitamins. Design: A randomized double-blind placebo controlled balanced incomplete Latin square design using ®ve spreads and four periods. The ®ve study spreads included butter, a commercially available spread and three experimental spreads forti®ed with three different concentrations of plant sterols. Subjects: One hundred apparently healthy normocholesterolaemic and mildly hypercholesterolaemic volunteers participated. Interventions: Each subject consumed four spreads, each for a period of 3.5 week. Results: Compared to the control spread, total cholesterol decreased by 0.26 (CI: 0.15 ± 0.36), 0.31 (CI: 0.20 ± 0.41) and 0.35 (CI: 0.25 ± 0.46) mmol=L, for daily consumption of 0.83, 1.61 and 3.24 g plant sterols, respectively. For LDL-cholesterol these decreases were 0.20 (CI: 0.10 ± 0.31), 0.26 (CI: 0.15 ± 0.36) and 0.30 (CI: 0.20 ± 0.41). Decreases in the LDL=HDL ratio were 0.13 (CI: 0.04 ± 0.22), 0.16 (CI: 0.07 ± 0.24) and 0.16 (CI: 0.07 ± 0.24) units, respectively. Differences in cholesterol reductions between the plant sterol doses consumed were not statistically signi®cant. Plasma vitamin K1 and 25-OH-vitamin D and lipid standardized plasma lycopene and alpha-tocopherol were not affected by consumption of plant sterol enriched spreads, but lipid standardized plasma (alpha beta)-carotene concentrations were decreased by about 11 and 19% by daily consumption of 0.83 and 3.24 g plant sterols in spread, respectively. Conclusions: The three relatively low dosages of plant sterols had a signi®cant cholesterol lowering effect ranging from 4.9 ± 6.8%, 6.7 ± 9.9% and 6.5 ± 7.9%, for total, LDL-cholesterol and the LDL=HDL cholesterol ratio, respectively, without substantially affecting lipid soluble (pro)vitamins. No signi®cant differences in cholesterol lowering effect between the three dosages of plant sterols could be detected. This study would support that consumption of about 1.6 g of plant sterols per day will bene®cally affect plasma cholesterol concentrations without seriously affecting plasma carotenoid concentrations. Sponsorship: Unilever Research Vlaardingen, NL Descriptors: cholesterol-lowering diet; plant sterols; dosage; spreads; lipid soluble (pro)vitamins

Introduction varying between 3 and 6% (Tang et al, 1998). Dietary factors contributing signi®cantly to a reduction of total One of the major modi®able risk factors for coronary heart cholesterol by dietary advice include high concentrations of disease (CHD) mortality is plasma cholesterol concentra- linoleic acid and low concentrations of saturated fatty acids tion. Several large scale intervention trials have shown that (Keys et al, 1965; Mensink & Katan, 1992; Katan et al, cholesterol lowering drug therapy in hypercholesterolaemic 1994). patients bene®cially affects CHD mortality risk (Gould et The cholesterol lowering properties of plant sterols have al, 1998). Mildly hypercholesterolaemic subjects usually been known since the 1950s (Pollak, 1952). High dosages have dietary guidelines prescribed as a ®rst treatment to of plant sterols have been applied to lower plasma choles- lower their blood cholesterol. Normocholesterolaemic sub- terol concentrations in hypercholesterolaemic subjects jects may apply such a diet as a part of a healthy life-style. (Ling & Jones, 1995). No obvious side effects have been However, the percentage reduction in blood total choles- observed in humans, except in individuals with phytoster- terol attributable to dietary advice is modestly effective in olemia (LuÈtjohann & Bergmann, 1997). The mechanisms free-living subjects; that is a reduction of total cholesterol of hypocholesterolaemic action may include inhibition of cholesterol absorption (Pollak & Kritchevsky, 1981). *Correspondence: Dr HFJ Hendriks, TNO Nutrition and Food Research Recently, it was suggested that substitution of sitostanol- Institute, PO Box 360, 3700 AJ Zeist, The Netherlands. Received 4 August 1998; revised 12 November 1998; accepted ester enriched spread for a portion of normal dietary spread 24 November 1998 is suitable as a strategy to reduce serum cholesterol in the Cholesterol lowering effect of plant sterols HFJ Hendriks et al 320 population. A one-year intervention study (Miettinen et al, apparently healthy. Health status was assessed by disease 1995) showed that consumption of sitostanol-ester enriched history questionnaire, physical examination and routine spread by subjects with mild hypercholesterolaemia was blood chemistry. Eligible subjects did not suffer from effective in lowering serum total cholesterol and LDL chronic gastrointestinal complaints and=or cardiovascular cholesterol by 10 and 14%, respectively. Sitostanol-ester, disease, high blood pressure (according to WHO guide- which is derived from hydrogenation of sitosterol from pine lines), high blood cholesterol (fasting serum total choles- tree woodpulp and subsequently esteri®ed with a free fatty terol < 7.5 mmol=L). They also did not use prescribed acid, is not the only effective cholesterol lowering plant medication (except oral contraceptives) and were not on a sterol. Sterols from commonly used edible oils (soybean, diet for medical reasons. Eligible subjects reported inten- rapeseed and sun¯ower oils), that is the 4-desmethylsterols sive exercise for not more than 10 h=week, consumption of sitosterol, campesterol and stigmasterol, have a similar a habitual Dutch diet including the use of spreads (assessed cholesterol lowering effect (Weststrate & Meijer, 1998). by dietary questionnaire) and consumption of alcoholic However, plant sterols, including stanols, may also beverages for less than 22 units=week when female and lower plasma concentrations of other lipophilic compounds for less than 29 units=week when male. as reported for carotene, lycopene and alpha-tocopherol (Weststrate & Meijer, 1998; Gylling et al, 1996). There- Experimental design fore, an optimal plant sterol or stanol intake level has to be The study had a double-blind, placebo-controlled, balanced selected, reducing cholesterol concentrations optimally, but incomplete Latin square design using ®ve spreads and four having minimal effect on other lipophilic compounds. periods. This design controls for between subject variation Relatively few studies, however, have evaluated the dose- and for variation over time. In addition this design results in effect relation of dietary plant sterols or stanols. Usually a balanced randomization of treatment orders over the one intake level has been applied in ef®cacy studies using subjects, so that any systematic effect of the order in mildly hypercholesterolaemic subjects. Sitostanol and which treatments were given will not create bias in the sitostanol ester dosages in the various studies ranged comparison of the treatment means. Subjects received in from 0.7 g=d (Miettinen & Vanhanen, 1994) to 3.4 g=d four consecutive periods of 24 or 25 d (3.5 weeks) 25 g (Pollak, 1952). Miettinen et al, 1995 compared 1.8 g=d spread per day. The included subjects were randomly and 2.6 g=d of sitostanol ester in the second part of his allocated to the spreads. One hundred volunteers were study showing a slightly stronger cholesterol lowering included and consequently data were obtained for each effect of the high dose on both total and LDL cholesterol. spread from eighty subjects. The ®ve spreads included Sitosterol has been consumed in dosages ranging from butter and a control spread, that is a commercially available 0.7 g=d (Miettinen & Vanhanen, 1994) to 6 g=d (Weisweiler spread Flora (Van den Bergh Foods, Crawley, UK) and et al, 1984) in hypercholesterolaemic subjects and soy bean three test spreads, which were spreads forti®ed with three oil sterols in dosages from 0.7 g=d (Pelletier et al, 1995) to concentrations of plant sterols, namely 3.37% (w=w), 3.0 g=d (Mensink & Katan, 1992) in normocholesterolae- 6.47% (w=w) and 13.06% (w=w) derived from commonly mic and mildly hypercholesterolaemic volunteers. used edible oils. These three test spreads should provide The main objective of this present study was to inves- 0.85 g=d, 1.62 g=d and 3.26 g=d of plant sterols, respec- tigate the dose-dependency of the cholesterol-lowering tively, assuming 100% compliance of the subjects to the effects of three relatively low dosages, namely 0.83, 1.61 treatments. Flora has a very similar fat composition as the and 3.24 g=d, of plant sterols derived from commonly used spreads containing plant sterols, and was used as the control edible oils in spreads. These three spreads, and a control spread for the quantitative evaluation of the cholesterol spread of similar fatty acid composition but not enriched in lowering effects of plant sterols. Butter was used as alter- plant sterols and butter were used by normocholesterolae- native, commonly used spread, with a completely different mic and mildly hypercholesterolaemic subjects consuming fat composition. The spreads were intended to replace an their habitual Dutch diet. In addition, the effects on lipid- equivalent amount of the spread(s) habitually used by the soluble (pro)vitamins including (alpha beta)-carotene, volunteers. To ensure all the spread was consumed, the lycopene, alpha-tocopherol, vitamin K1 and 25-OH-vitamin volunteers were instructed not to use it for cooking or D were evaluated. frying. Half a portion was to be used with lunch and the other half with dinner, for example by mixing it with the meal on the subjects' plate or by spreading it on bread or Subjects and methods toast eaten at dinner. Volunteers and TNO personnel were The study was conducted according to Good Clinical blinded. Unilever packed all test spreads in tubs of 25 g and Practice at the TNO Nutrition and Food Research Institute, delivered them in boxes with a blind code. The tubs were Zeist, Netherlands. The study protocol was approved by the subsequently labelled with the subject number and study TNO Medical Ethical Committee. period. Spreads were dispensed to the subjects twice per spread Subjects period. The volunteers received a lea¯et containing instruc- Subjects were recruited from a pool of volunteers contain- tions on storage conditions and consumption of the spreads, ing no TNO employees. Respondents received a verbal containing also space for the speci®cation of deviations brie®ng and received the same information in writing. They from the instructions. Each subject was provided with one signed for informed consent and ®lled in a questionnaire on portion for each study day and four spare portions per study life style, disease history and dietary habits. Each of the period in case of damage or loss. Portions not consumed respondents was physically examined and blood was col- (including spare portions) were returned to TNO halfway lected after an overnight fast, for routine blood chemistry. and at the end of the period. Compliance was calculated A total of 147 subjects was screened of which 100 based on subjects speci®cation of deviations and on the were included in the study. All volunteers included were number of portions returned. In case of con¯icting Cholesterol lowering effect of plant sterols HFJ Hendriks et al 321 information, the worst case was assumed. During the study, less to butter as compared to Flora; oleic acid contents are subjects maintained their daily routines and used a self- 15.1 and 20.1 g=100 g, respectively and the sum of linoleic selected diet. and linolenic acid content are 1.5 and 33.0 g=100 g, respec- Fasting blood was sampled at the end of each period tively. In addition, butter has a higher content of trans fatty (after 3.5 weeks). Analyses were carried out after ending acids as compared to Flora (2.9 and 0.4 g=100 g, respec- the clinical part of the study when all blood samples were tively). The experimental spreads have a trans fatty acid collected. Body weight was measured half way and at the content of 0.8 g=100 g. end of each period. Also, health status and medicine use were registered by questionnaire half way and at the end of Blood lipids and routine blood chemistry each period. Adverse events were reported by the volun- Blood was collected after an overnight fast from the teers by ®lling in a questionnaire halfway and at the end of antecubital vein using Vacutainer1 tubes. For serum collec- each period. The medical investigator consulted the volun- tion, blood was collected in tubes containing clot activator. teers when additional information was needed. Dietary Blood was centrifuged within one hour after collection and intake was assessed with a food frequency questionnaire serum stored at 7 18C. Alkaline phosphatase, ALP; at the end of each period. alanine transaminase, ALT; aspartate transminase, AST; gamma-glutamate transaminase, gamma-GT; total choles- Spreads terol; HDL-cholesterol, after precipitation with polyethy- The plant sterol enriched spreads were prepared by Van den lene glycol; and triglycerides were determined using Bergh Foods, Pur¯eet, UK. The spreads were forti®ed with commercial test kits (Boehringer, Mannheim, Germany) plant sterol (-ester) concentrates derived from vegetable on a Hitachi 911 automatic analyser (Hitachi Instrument (predominantly soybean) oil distillates (Henkel Corpora- Division, Ibaraki-ken, Japan). Serum LDL-cholesterol con- tion, LaGrange, USA). The vegetable oil sterols were centration was calculated using the formula by Friedewald esteri®ed with fatty acids from sun¯ower-seed oil to an et al, 1972. esteri®cation degree of 82% (Unilever Research Labora- tory, Vlaardingen, The Netherlands). The sterol concen- Plasma lipid-soluble (pro)vitamins trates were re®ned and used in spread production together Blood was collected in ice-chilled tubes containing lithium with other edible oils and fats (sun¯ower-seed oil, rapeseed heparin and put in a cool dark box immediately after oil and hard stock) to produce spreads as close as possible collection. Blood was centrifuged within half an hour in fatty acid composition as the non-forti®ed control. Total after collection, and plasma stored at 7 70C until ana- fat content, fatty acid composition, free sterols, carotenoids lyses. Lycopene, (alpha beta)-carotene, and alpha-toco- and alpha-tocopherol content of spreads were measured as pherol were determined by Unilever Research described previously (Weststrate & Meijer, 1998). (Vlaardingen, Netherlands) in all samples except for The butter and spread composition is given in Table 1. those obtained after butter consumption. Plasma was As expected, the plant sterol enriched spreads have a fatty extracted by addition of ethanol and n-heptane containing acid composition very similar to that of the control spread ethyl-beta-apo-80-carotenoate and alpha-tocopheryl-acetate (Flora), while the fatty acid composition of butter is as internal standards, followed by a second extraction step different. Butter contains more saturated fatty acids as with n-heptane=diethyl ether (1:1 (v=v)). A normal phase compared to Flora (59.9 and 15.3 g=100 g, respectively). Nucleosil 5N(CH3)2 column (Machery & Nagel, Dueren, As a consequence, unsaturated fatty acids contribute much Germany) was used with a ¯ow of 0.8 ml=min using

Table 1 Composition of spreads

3.37% (w=w) 6.47% (w=w) 13.06% (w=w) Component Butter Floraa plant sterols plant sterols plant sterols

Total fat as fatty acids (g=100 g) 83.6 69.4 70 70.6 69.8 Major fatty acids (g=100 g) Total Saturated fatty acids 59.9 15.3 15.3 15.7 15.7 Lauric (C12:0) 3.9 1.7 1.6 1.7 1.7 Myristic (C14:0) 10.7 0.7 0.7 0.7 0.7 Palmitic (C16:0) 28.4 9.3 9.3 9.5 9.6 Stearic (C18:0) 7.8 2.5 2.7 2.7 2.6 Oleic (C18:1 cis) 15.1 20.1 19.6 20 20.3 Linoleic (C18:2, 9c, 12c) 1.1 31.5 32.2 32.2 31 Linolenic (C18:3 9c, 12c, 15c) 0.4 1.5 1.4 1.3 1.4 Total trans 2.9 0.4 0.8 0.8 0.8 Total sterols (mg=100 g) 221 298 3370 6466 13057 Major sterols (mg=100 g) Cholesterol 213 2.9 14.2 27.6 111 Brassicasterol 13.9 56.4 101 204 Campesterol 2.2 63.3 825 1597 3290 Campestanol 4 25.4 82.7 122 Stigmasterol 12.6 602 1198 2494 Beta-sitosterol 1.9 159 1629 3044 6233 Sitostanol ± 56 120 162 Delta-5-Avenasterol 9.1 46 96.5 137 Delta-7-Stigmastenol 16.1 27.8 35.9 60.2 Alpha-tocopherol (mg=100 g) 1.8 31.7 32.7 30.1 26.3 (Alpha beta)-Carotene (mg=100 g) 0.6 0.5 0.7 0.8 0.8

aControl: not enriched in sterols. Cholesterol lowering effect of plant sterols HFJ Hendriks et al 322 n-heptane as the mobile phase. The carotenoids were for all subjects; 3 for 60 subjects and 4 for 20 subjects. detected at 450 nm, alpha-tocopherol was detected at Therefore, analysis was carried out with a simpli®ed model 294 nm (Weststrate & Van't Hof, 1995). not including residual effects and for n 60 subjects. Plasma carotenoid and alpha-tocopherol concentrations Data are expressed as the meanÆ s.d. Presented changes were standardised for plasma lipid (total cholesterol total and their 95% con®dence interval (CI) are calculated using triacylglycerol) concentrations, because the sterol enriched the least square means. Data which were not distributed spreads affected the concentrations of blood lipoproteins, normally were ln transformed before statistical analysis which are the plasma carriers of carotenoids and alpha- (only (alpha beta)-carotene and vitamin K1). The percen- tocopherol. tal change and the 95% con®dence intervals calculated Vitamin K1 and 25-OH-vitamin D were analysed at the from the transformed data were transformed again to the TNO Institute in all samples except for those obtained after original scale in order to facilitate interpretation. The least butter consumption. For vitamin K1 analyses plasma was signi®cant difference between spreads is expressed as a extracted by addition of methanol and hexane and 20,30- percentage (last column of Table 5). The null hypothesis dihydrophylloquinone was added as internal standard. was rejected at the 0.05 level of probability. The statistical Vitamin K1 was analysed by HPLC, using an Inertsil 5 analyses were performed using Genstat 5 Release 3.1 ODS-3 reversed-phase column (Chrompack) with post (Lawes Agricultural Trust, Rothamsted Experimental column electrochemical reduction and ¯uorimetric detec- Station). tion (Haard et al, 1986). For vitamin D analyses plasma The study was unblinded after reporting of the adverse was extracted with a mixture of dichloromethane, methanol events, evaluation and statistical analysis of food intake and water, and 25-OH-vitamin D was isolated by chroma- data and all biochemical data except vitamin D and K. tography on a small silica column. The 25-OH-vitamin D content was analysed using a competitive protein binding Results assay (Berg et al, 1991). General Dietary intake The study started with 100 volunteers, 42 males and 58 Dietary intake, excluding the spreads, was assessed at the females, with a mean age of 37Æ 10 y ranging from 19 ± end of each of the four periods using a modi®ed version of 58 y, and a mean body mass index of 22.8Æ 2.5 kg=m2 the validated TNO food frequency questionnaire (Grooten- ranging from 17.7 ± 28.6 kg=m2. Baseline fasting total-, huis et al, 1995). The questionnaire was applied to assess LDL- and HDL-cholesterol concentrations were 5.10Æ continuity in the intake of macronutrients, main fats, total 0.97 mmol=L (range: 2.71 ± 7.42), 2.97Æ 0.83 mmol=L energy, alcohol and ®bre. The questionnaire was modi®ed (range: 1.12 ± 5.22), and 1.65Æ 0.38 mmol=L (range: 0.75 ± to obtain more detailed information on the intake of 2.63), respectively. All participants completed the study. speci®c brands of spreads and dietary fats in the habitual Compliance was generally very good, only about 1% of diet and consequently to better evaluate fat composition the portions of the spreads was not consumed, distributed (like SAFA, MUFA, PUFA) of the habitual diet during equally over the ®ve spreads. Average spread consumption each period. Subjects completed the questionnaire during was 24.7 g=d for the 3.37% (w=w) plant sterol spread, their visit to the Institute at the last day of each period. The 24.9 g=d for the 6.47% (w=w) plant sterol spread and dietitian then checked the questionnaires for completeness, 24.8 g=d for the other three spreads. Consequently, the and, where necessary, asked subjects to check questions at average daily plant sterol consumption was 0.83 g, 1.61 g home and return answers by mail. For calculation of the and 3.24 g, respectively. nutrient intake the `Netherlands Food Composition Table Body weights were 72.3Æ 11.0, 71.4Æ 10.9, 72.1Æ (NEVO)' was used. Speci®c fat composition data were 11.2, 72.2Æ 11.2 and 72.0Æ 10.4 kg after butter and Flora expanded upon a large number of extra brands of spreads consumption and after daily consumption of 0.83, 1.61 and and dietary fats, as provided by Unilever. 3.24 g plant sterols, respectively. Body weights differed only slightly after daily consumption of 1.61 and 3.24 g Statistical analyses plant sterols, the difference being 0.3 kg (P < 0.01). This Differences in variables measured between spreads were small difference is not expected to have affected the out- evaluated by a two-sided ANOVA, using the following come of this study. factors: subject, spread, period and residual effects (includ- Nutrient intake, excluding spread intake, during the four ing carry-over). Two analyses were performed: residual periods is shown in Table 2. Total fat intake and the effects corrected for direct effects and direct effects cor- contribution of saturated, monounsaturated and polyunsa- rected for residual effects. Residual effects were observed turated fatty acids to fat intake, dietary cholesterol and for total and LDL-cholesterol and consequently the direct energy intake did not change during consumption of effects were corrected for these residual effects. Residual spreads, nor did any other calculated dietary intake effects observed were small, and correction for them did parameter. not affect the results. In this present study residual effects No side effects occurred after the consumption of any of may well be a consequence of the design and may not have the spreads applied in these concentrations and under these a biological basis. conditions. Side effect testing included the liver enzymes Since the lipid-soluble (pro)vitamins were not analysed ALP, ALT, AST and gamma-GT (Table 3) and adverse after butter consumption, for those parameters the above events reporting. mentioned design could not be used. Correction for residual effects was not possible because data for the previous Blood lipids period were incomplete, that is, absent when butter was Total, LDL-, and the LDL=HDL cholesterol ratio were all consumed in the previous period. In addition, the number of decreased by plant sterol consumption, while triacylgly- spreads (for which data were available) was not the same cerol concentration was not affected (Table 4). Cholesterol lowering effect of plant sterols HFJ Hendriks et al 323 Table 2 Intake of energy, protein, fats, carbohydrates, dietary ®bre and alcohol as assessed by a food frequency questionnaire at the end of each treatment period, covering that treatment period. Intakes are expressed as grammes and as percent energy (en%) during consumption of butter, Flora and 0.83, 1.61 and 3.24 g plant sterols. Values do not include intake of spreadsa

Butter Flora 0.83 g plant sterols 1.61 g plant sterols 3.24 g plant sterols Nutrients (n 79) (n 79) (n 79) (n 79) (n 80)

Energy (kJ) 9536Æ 2715 9432Æ 2381 9270Æ 3196 9304Æ 2752 9762Æ 3243 Protein (g) total 93.1Æ 26.7 91.9Æ 24.6 91.6Æ 35.5 91.2Æ 29.0 94.0Æ 31.6 vegetable 33.8Æ 13.3 33.0Æ 9.8 32.8Æ 12.2 32.8Æ 11.6 34.8Æ 14.5 Fat (g) total 87.5Æ 33.0 87.1Æ 28.6 83.1Æ 32.8 85.1Æ 31.8 90.5Æ 35.3 SAFAb 34.9Æ 13.5 34.5Æ 11.7 33.6Æ 13.3 33.9Æ 12.9 35.7Æ 14.4 MUFAb 30.2Æ 11.6 29.9Æ 10.2 28.2Æ 10.8 29.2Æ 11.1 31.6Æ 12.8 PUFAb 15.7Æ 7.8 16.0Æ 6.6 14.6Æ 6.9 15.2Æ 7.4 16.4Æ 8.1 cholesterol (mg) 254.0Æ 94.9 244.8Æ 83.8 249.5Æ 166.9 248.7Æ 100.2 257.1Æ 100.3 Carbohydrates (g) total 261.4Æ 79.8 256.4Æ 71.0 259.2Æ 94.3 254.5Æ 77.3 267.4Æ 96.2 mono=-disaccharides 124.5Æ 43.6 124.5Æ 46.2 124.6Æ 54.8 124.2Æ 44.1 128.4Æ 51.3 polysaccharides 136.8Æ 47.4 131.8Æ 39.1 134.6Æ 52.3 130.3Æ 45.3 139.1Æ 54.2 Dietary ®bre (g) 26.2Æ 9.2 25.7Æ 6.9 25.7Æ 8.6 25.8Æ 8.2 26.3Æ 9.5 Alcohol (g) 9Æ 11 10Æ 13 8Æ 9 10Æ 12 9Æ 11 Protein (en%) total 16.9Æ 2.9 16.8Æ 2.7 17.0Æ 2.8 16.8Æ 2.7 16.6Æ 2.2 vegetable 6.0Æ 1.4 6.0Æ 1.1 6.1Æ 1.2 6.0Æ 1.1 6.0Æ 1.1 Fat (en%) total 33.5Æ 6.3 33.9Æ 5.6 32.9Æ 5.2 33.3Æ 5.5 33.9Æ 5.5 SAFA 13.5Æ 3.3 13.5Æ 2.9 13.4Æ 2.5 13.3Æ 2.7 13.5Æ 2.86 MUFA 11.6Æ 2.3 11.6Æ 2.1 11.2Æ 2.0 11.4Æ 2.0 11.8Æ 2.2 PUFA 5.9Æ 1.8 6.2Æ 1.6 5.7Æ 1.7 5.9Æ 1.7 6.0Æ 1.7 cholesterol=energy (mg=MJ) 26.9Æ 7.2 26.0Æ 6.2 26.4Æ 7.4 26.7Æ 6.6 26.7Æ 6.5 Carbohydrates (en%) total 46.8Æ 7.2 46.3Æ 6.3 47.6Æ 6.3 26.9Æ 6.5 46.7Æ 6.7 mono-=disaccharides 22.3Æ 5.6 22.4Æ 6.0 22.8Æ 6.2 23.0Æ 5.8 22.5Æ 5.8 polysaccharides 24.5Æ 5.0 23.9Æ 4.5 24.8Æ 4.9 23.9Æ 4.7 24.2Æ 4.0 Alcohol (en%) 2.8Æ 3.2 3.1Æ 3.8 2.6Æ 2.6 3.1Æ 4.0 2.8Æ 3.7 Dietary ®bre=Energy (g=MJ) 2.8Æ 0.8 2.8Æ 0.8 2.9Æ 0.9 2.9Æ 1.0 2.8Æ 0.7 aButter contained 20.9 g fat and spreads contained 17.5 g fat, and contributed for about 8 and 7% of energy intake, not included in this table. bSAFA, saturated fatty acids; MUFA, mono-unsaturated fatty acids; PUFA, poly-unsaturated fatty acids.

Table 3 ALP, ALT, AST and g-GT concentrations after consumption of butter, Flora and 0.83, 1.61 and 3.24 g plant sterolsa,b

Butter Flora 0.83 g plant sterols 1.61 g plant sterol 3.24 g plant sterol (n 80) (n 80) (n 80) (n 80) (n 80)

ALP (U=L) 60Æ 14 59Æ 14 58Æ 15 58Æ 15 59Æ 15 ALT (U=L) 15Æ 6 15Æ 6 15Æ 7 15Æ 6 15Æ 8 AST (U=L) 19Æ 5 19Æ 5 19Æ 5 20Æ 5 20Æ 6 g-GT (U=L) 18.2Æ 9.3 17.6Æ 8.7 17.8Æ 9.4 18.2Æ 9.6 18.1Æ 9.3

aALP alkaline phosphatase; ALT alanine transaminase; AST aspartate transaminase; Gamma-GT gamma-glutamate transaminase. bNo statistical signi®cant differences between treatments were present.

Total and LDL-cholesterol concentrations decreased HDL-cholesterol was not affected after consumption of after the consumption of plant sterol enriched spread the plant sterol enriched spreads as compared to Flora as compared to Flora (P < 0.001) and butter (P < consumption. However, HDL-cholesterol was decreased 0.001). Decreases in total cholesterol as compared to after consumption of 1.61 and 3.24 g plant sterols as Flora were 0.26 (CI: 0.15 ± 0.36), 0.31 (CI: 0.20 ± 0.41) compared to butter consumption (P < 0.01). Decreases and 0.35 (CI: 0.25 ± 0.46) mmol=L, for daily consumption were 0.047 (CI: 0.013 ± 0.081) and 0.051 (CI: 0.017 ± of 0.83, 1.61 and 3.24 g plant sterols, respectively. For 0.085) mmol=L, respectively. LDL-cholesterol these decreases were 0.20 (CI: 0.10 ± Analysis of variance for the LDL=HDL cholesterol ratio 0.31), 0.26 (CI: 0.15 ± 0.36) and 0.30 (CI: 0.20 ± shows a pattern very similar to that of total cholesterol and 0.41) mmol=L, respectively. Differences in cholesterol LDL-cholesterol. The LDL=HDL ratio decreased after the reductions between the plant sterol doses consumed were consumption of plant sterol enriched spreads as compared not signi®cant. Total and LDL-cholesterol concentrations to Flora and as compared to butter. Decreases as compared decreased after Flora consumption as compared to butter to Flora were 0.13 (CI: 0.04 ± 0.22), 0.16 (CI: 0.07 ± 0.24) consumption. and 0.16 (CI: 0.07 ± 0.24) units, after 0.83, 1.61 and 3.24 g Cholesterol lowering effect of plant sterols HFJ Hendriks et al 324 plant sterols, respectively. The LDL=HDL cholesterol ratio as a for Flora and butter were the same. B CI 0.04 0.03 0.02 0.01 0.09 ) 2.6%) 4.0%) 1.6%) 2.1%) ± 1.0%) ± 7 7 7 nces: (% 7 7 7 7 7 ± ± Plasma lipid-soluble (pro)vitamins ± ( ( ( ( ( 0.13 Plasma concentrations of vitamin K1 and 25-OH-vitamin D 0.11 7 7 substa 0.24 0.22 0.06 0.14 0.12 0.03 0.04 0.01

were not affected by consumption of the spreads enriched ecrease 7 7 7 80) 7 7 7 7 7 D

in plant sterols. The compounds (alpha beta)-carotene, study Butter n lycopene and alpha-tocopherol were decreased after con- ( C2 C2

sumption of plant sterol enriched spreads as compared to s.d. 0.94 0.83 0.39 0.82 0.43 Æ Flora consumption. Plasma carotenoid and alpha-toco- between Æ Æ Æ Æ Æ pherol data were also lipid standardized to correct for the an ence Me 5.27 3.15 1.65 2.02 decreases in plasma lipids (Table 5). 1.10 r

(Alpha beta)-carotene concentrations were decreased differ a CI

with 12%, 11% and 19% after consumption of 0.83, 1.61 Butte ant ) 0.46 0.41 0.24 0.06 and 3.24 g plant sterols, respectively, as compared to 0.15 (% ± ± ± ± ± and (6.8%) (9.9%) (1.5%) (7.8%) (4.8%)

consumption of Flora. The decrease in (alpha beta)-car- s ls signi®c 0.25 0.20 0.07 0.01 0.05 a 0.35 0.30 0.02 0.16 otene concentrations was larger after consumption of 3.24 g 0.05 e sterol 7 7 stero plant sterols as compared to consumption of 0.83 and 1.61 g ecrease D 80)

plant sterols (P < 0.05). (Alpha beta)-carotene concentra- plant plant n indicat ( g tions per total plasma lipids were also decreased, namely by g B4,C4 B4,C4 B3 B4,C4

about 8 and 15% after consumption of 0.83 and 3.24 g plant s.d. cters 3.24 3.24 Æ 0.93 0.79 0.40 0.89 sterols, respectively. 0.61 and chara Æ Æ Æ Æ Lycopene concentrations decreased by about 11 ± 15% Æ Mean 1.61 4.81 2.75 1.61 1.86 after consumption of spreads enriched with plant sterols, 1.08 respectively. No differences between the dose levels of script a plant sterols were observed. However, plasma lycopene 0.83, CI Super

concentrations per total lipid in plasma was not affected by 1. 0.41 0.36 0.06 0.24 0.17 .9%) .5%) .3%) .9%) .1%) se. (%) ± ± ± ± ± Flora, (5 (8 (1 (7 consumption of any of the spreads enriched in plant sterols (6 0.00 ls of ase < 0.20 0.15 0.01 0.07 (Table 5). 0.03 decrea 0.31 0.26 0.02 0.16 0.07 P on 4 7 7 stero

Alpha-tocopherol concentrations were decreased by ; the Decre 80) nt about 6 and 8% after consumption of 1.61 and 3.24 g mpti of 0.01 pla n ( plant sterols, respectively. The highest plant sterol concen- < g val consu P 3 r

tration decreased alpha-tocopherol concentrations further B4,C4 B4,C4 B3 B4,C4 ; s.d. 1.61 inter Æ as compared to the lowest concentration. However, alpha- afte 0.92 0.76 0.80 0.47 0.41 0.05 Æ Æ Æ Æ tocopherol concentrations per total lipid in plasma were not Æ < dence Mean affected by consumption of any of the spreads enriched in P 2 rations 4.84 2.77 1.84 1.05 plant sterols. 1.63 con® nce: a and CI concent ) 0.36 0.31 0.04 0.22 Discussion 0.19 (%) (% ± ± ± ± ± signi®ca (4.9%) (6.7%) (6.5%) (8.4%) (0.3%) e

In this double-blind, placebo-controlled trial in normocho- of 0.15 0.10 0.03 0.04 0.00 lglycerol valu 0.26 0.20 0.13 0.09 0.01

lesterolaemic and mildly hypercholesterolaemic adult crease 7 7 sterols level ute De 80)

volunteers no statistical signi®cant dose-dependency was triacy , the plant

observed in the cholesterol lowering effects of three dif- n e ( absol g ratio ferent dosages of plant sterols. Intake of total plant sterols ,C4 ,C4 ,C4 B4 B4 B3 s.d.

as low as 0.83 g=d, already resulted in signi®cant decreases 0.83 indicat Æ sterol 0.91 0.48 0.41 0.78 0.82 of blood cholesterol. Daily intake with lunch and dinner of mption, Æ Æ Æ Æ Æ

plant sterol enriched spreads decreased total cholesterol, Mean chole ®gures consu 4.94 1.03 1.65 2.86 1.87

LDL cholesterol and the LDL=HDL ratio, but did not affect L

plasma triacylglycerol and HDL cholesterol concentrations. HD = Flora rscript

Decreases in total cholesterol were 4.9 ± 6.8%, decreases in s.d. to 0.94 0.62 0.40 0.85 0.91 80) Æ LDL d Æ Æ Æ Æ Æ Supe

LDL cholesterol 6.7 ± 9.9% and decreases in LDL=HDL , Flora n a. (

ratio were 6.5 ± 7.9% as compared to control spread con- pare 5.16 1.13 1.64 3.05 2.01 Mean

sumption. The reductions in total and LDL-cholesterol Flor com

observed are large and may, on a population basis, sub- to as cholesterol s

stantially contribute to the prevention of coronary heart L esterol

disease (Law et al, 1994). HD mpared mean L) L) L) L) co e = =

Theoretically, factors other than spread consumption = = chol and ol as

may have affected the outcome of the study. However, C squar (mmol HDL (mmol (mm LDL

the study was double-blinded and compliance appeared to (mmol =

be extremely good. Also, intake of nutrients including least ols butter, eter of to Total,

dietary fats and total energy, did not differ signi®cantly esterol esterol esterol esterol

during consumption of the spreads as assessed by food 4 red lglycer chol param chol chol chol frequency questionnaire. Moreover, consumption of plant e

sterol enriched spreads appeared to have no adverse side Decrease compa Lipid Triacy LDL HDL Tabl Total a LDL Cholesterol lowering effect of plant sterols HFJ Hendriks et al 325 effects, de®ned as several liver enzyme concentrations and

(%) adverse events reporting. In this present study a mixture of plant sterols obtained ence from commonly used edible oils (predominantly soybean)

differ was applied consisting mainly of sitosterol also containing 3 23 3.6 7.8 7.6 18.5 10.6 10.6 appreciable amounts of campesterol and stigmasterol. This ant mixture resulted in reductions of total and LDL cholesterol of 4.9 ± 6.8% and 6.7 ± 9.9%, respectively. These reductions signi®c

t are slightly lower as those reported in literature on both

Leas soybean derived plant sterols and wood pulp derived sitostanol esters. Pelletier et al, 1995 applied soybean plant sterols and reported slightly higher reductions in 4 e total and LDL-cholesterol (10 and 15%, respectively). In 2.8 7.8 0.6 19.2 15.4 14.9 10.2 18.6 most other studies the plant sterols applied were sitostanol ls 7 7 chang 7 7 7 7 7 ester (Miettinen et al, 1995; Vanhanen et al, 1993; Vanha- % stero nen et al, 1994). The study by Miettinen et al (1995), nt treating 153 subjects with a mild hypercholesterolaemia, pla c3 b2 c3 g

c3 resulted in a 10% reduction of total cholesterol and a 14% 7 s.d. 5.3 0.05 26 0.28 0.19 0.03 0.61 112 Æ reduction of LDL cholesterol after 12 months in the study. 3.24 Æ Æ Æ Æ Æ Æ Æ Æ After three months of sitostanol consumption, however, the 70 168 Mean 25.6 0.07 0.40 0.37 0.06 4.47 reduction of total cholesterol was approximately 6.5%, which is similar to the reduction reported in this present study. Vanhanen et al, reported reductions of LDL choles- 4 e 7 terol of 10 ± 15% as compared to control in two studies with 5.9 5.5 7.9 0.3 7.4 11.0 12.8

ls hypercholesterolaemic subjects (Vanhanen et al, 1993, 7 7 7 7 7 chang 7 7 60)

% 1994) and Heinemann et al, 1986 report a cholesterol stero

n reduction of 15% in severe hypercholesterolaemic subjects nt ( s as well. Weststrate & Meijer (1998) compared spreads pla b2 b1 ab g bc2

6 enriched in plant sterols derived from commonly used s.d. sterol 0.51 0.22 5.2 0.11 0.04 0.65 27 136 Æ t edible oils and enriched in sitostanol ester. Reductions of 1.61 Æ Æ Æ Æ Æ Æ Æ Æ plasma total and LDL cholesterol in their study were 8 and plan 73 206 Mean g 0.52 0.36 25.8 0.10 0.06 4.51 13%, respectively, independent of the plant sterols applied. 0.001. These reductions correspond to the reductions in total and 3.24 < P d

4 LDL cholesterol reported in this study. , 3 an In this present study the effects of the three dosages of 3.0 8.2 7.1 2.1 5.4 4.1 11.7 11.0 7 7 7 7 1.61 7

change plant sterols did not show a clear difference in ef®cacy. 0.01; 7 7 , ). < % However, 95% con®dence intervals (as compared to con- sterols P 0.83 0.05 ,

2 trol) suggest increasing cholesterol reductions with increas- < and P plant ing plant sterol content (Table 4). Dose-effect relationship ( , b2 b1 b1 g ab 0.05;

s.d. was further investigated by analysing the signi®cance of a < 0.47 0.21 5.1 0.09 0.03 0.83 148 26 Æ Flora 0.83 P trend in the three dose levels of plant sterols, using their Æ Æ Æ Æ Æ Æ Æ Æ , an 1 of

different orthogonal polynomials. This trend may either be linear or 71 198 Me 0.50 0.38 26.9 0.09 0.06 4.63 ly quadratic, because only three dose levels were included. ora: tion Fl

mp The analysis, however, did not show a linear (P 0.076 for to

a a a total cholesterol) nor a quadratic (P 0.972 for total a ed s.d. consu signi®cant 0.47

5.3 cholesterol) trend in any of the serum lipids analysed. 0.09 0.22 0.03 0.72 167 25 Æ Æ Æ Æ Æ Æ Æ Æ Æ mption.

are The statistical power of the study as executed indicated Flora after 68 compar

228 that an overall signi®cant difference between plant sterol Mean 0.57 28.1 0.10 0.40 0.07 4.59 ns row as consu a

a concentrations, at a probability of P 0.05 and expressed

ance as a percentage of the grand mean, was 2.2 and 3.7% for ntratio Flor within

to total and LDL cholesterol, respectively. This corresponds d conce

mmol) to a difference of 0.11 mmol=L for both parameters. The signi®c = ters ) pare

of overall effect on total cholesterol varied between 0.15 and mol m lycerol. 0.46 mmol=L with the average decrease for the three ( com charac vitamin mmol 5 = level as pt concentrations of plant sterols being about 0.05 mmol=L s (pro) L) TG) mol the higher with increasing dose, namely 0.26, 0.31 and = triacylg m l e mmol) e = 0.35 mmol=L for 0.83, 1.61 and 3.24 g plant sterol con- mean mo L) L) m superscri TG)( e (TC mol ( = TG = sumption. Therefore, the statistical power of the study may = m -solubl ; indicat ( s ) rent not have been suf®cient to detect the small differences in mol squar m (nmol (TC lipid otene otene ( L) TG) =

mL effect between the three plant sterol concentrations. = a esterol = D diffe l ®gure

least Information on the dose dependency for blood lipid in C mo (pg eter herol herol chol ta)-car ta)-car of m (T with ( Plasm l lowering by plant sterols is scarce. Only one experiment = be be K1 erscript

5 has directly compared the ef®cacy of different doses of -vitam tota ene ene param a a a-tocop a-tocop 59. 58. min

change plant sterols. In Miettinen's long-term experiment (Mietti- Sup Values ble 3 c ± ± % TC n n nen et al, 1995) with a dose of 2.6 g sitostanol ester per day (Alph Alph (Alph Lycop Lycop Ta a 1 4 5 6 7 Alph Vita Lipid 25-OH Cholesterol lowering effect of plant sterols HFJ Hendriks et al 326 for 12 months, half of the subjects received a dose of 1.8 g positive effects on health (Poppel, 1993; Manson et al, per day during the last six months. At the end of the study 1993). Minimization or compensation for the carotenoid the difference in cholesterol reduction between the two lowering effects of plant sterol enriched spreads should be dosages was only 0.2 mmol=L. This suggests that concen- considered. tration dependency of the cholesterol lowering effect may All three doses of plant sterols were effective in sub- be shallow in this concentration range. Miettinen et al, stantially reducing total and LDL cholesterol. There may or 1995 suggest that the practical interpretation is that these may not be a dose dependency. Our study shows that both two doses produce a similar cholesterol-lowering effect. the lowest and the highest dose, but not the middle dose Dose-dependency may not only be shallow, some indi- affect plasma carotenoid concentrations to a limited extent. cations for a threshold dosage are reported. Miettinen et al, Therefore, this study would support that consumption of 1995 reported small reductions of cholesterol (2.4% as about 1.6 g of plant sterols per day will bene®cially affect compared to an increase of 1.9% in control group) after plasma cholesterol concentrations without seriously affect- consumption of 0.7 g plant sterols per day. VanHanen et al ing plasma carotenoid concentrations. reported a non signi®cant reduction of cholesterol (7%) after intake of 0.8 g plant sterols per day. These experi- ments, however, were performed with sitostanol ester Conclusions derived from wood pulp, but not with plant sterols derived This present study shows that daily intake of spreads from commonly used edible oils. The experiment presented enriched with 0.83 ± 3.24 g plant sterols derived from com- here showed a substantial cholesterol lowering effect of monly used edible oils decreases total cholesterol by 0.15 ± 0.83 g plant sterols. Also Pelletier et al, 1995 reported a 0.46 mmol=L as compared to consumption of control signi®cant cholesterol reduction using a low dose, namely spread. LDL-cholesterol is decreased by 0.10 ± 0.41 0.74 g, soybean plant sterols. These data suggest that, at mmol=L as compared to control. LDL=HDL ratio is least for these plant sterols, a low dose is effective. decreased by 0.04 ± 0.24 as compared to control. The test One hypothesis put forward to explain the absence of a spreads do not affect HDL cholesterol as compared to dose-dependency is that a compensatory increase in cho- control spread, but the spreads with the highest plant lesterol synthesis occurs after consumption of higher sterol contents do slightly reduce HDL cholesterol as dosages of plant sterols. VanHanen et al (1993) evaluated compared to butter. Lipid standardised plasma alpha- concentrations of precursors of cholesterol and estimated tocopherol and lycopene were not affected by the plant that consumption of 2 g sitostanol per day (but not 0.8 g=d) sterol enriched spreads. However, lipid standardized by hypercholesterolaemic subjects increased cholesterol plasma (alpha beta)-carotene was reduced by daily con- synthesis by 2 mg per day per kg body weight. Combining sumption of 0.83 and 3.24 g plant sterols. this with reported inhibitions of 30% (Vanhanen et al, 1994) to 60% (Gylling & Miettinen, 1994) in the absorption Acknowledgements ÐThe authors acknowledge the contribution of all of cholesterol originating from both the diet as well as from those involved in the conduct of the study at the TNO Nutrition and the enterohepatic circulation, leads to the conclusion that Food Research Institute. We also thank the volunteers for their enthusiastic increased cholesterol synthesis may at least in part com- participation. pensate for a reduction in cholesterol absorption. Since the reduction of cholesterol by plant sterols is thought to be mainly due to inhibition of cholesterol absorp- References tion, plant sterols might also interfere with the absorption of Berg H van den, Schrijver J & Boshuis PG (1991): Vitamin D (25-OHD) in other fat-soluble compounds like the fat-soluble vitamins. serum by competitive protein-binding assay. In: Nutritional Status To the best of our knowledge, the effects of plant sterol Assessment. A Manual for Population Studies F Fidanza, (ed). intake on the status of vitamins D and K has not been London: Chapman & Hall, pp 203 ± 209. reported on. In this present study no effects of any of the Friedewald WT, Levy RI & Frederickson DS (1972): Estimation of the spreads enriched in plant sterols were observed. In a long- concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin. Chem. 18, 499 ± 502. term study using a sitostanol-ester enriched spread (Gylling Gould AL, Rossouw JE, Santanello NC, Heyse JF & Furberg CD (1998): et al, 1996) alpha-tocopherol and carotene concentrations Cholesterol reduction yields clinical bene®ts: impact of statin trials. were decreased after consumption of 2.6 g sitostanol ester Circulation 97, 946 ± 952. per day. Lipid standardized alpha-tocopherol concentration Grootenhuis PA, Westenbrink S, Sie CMTL, Neeling JND de, Kok FJ & Bouter LM (1995): A semiquantitative food frequency questionnaire for did, however, not change. In a previous study with spread use in epidemiologic research among the elderly: Validation by com- enriched in plant sterols derived from commonly used edible parison with dietary history. J. Clin. Epidemiol. 48, 859 ± 868. oils applying an intake of 3 g plant sterols per day (West- Gylling HK & Miettinen TA (1994): Serum cholesterol and cholesterol strate & Meijer, 1998), plasma (alpha beta)-carotene and and lipoprotein metabolism in hypercholesterolemic NIDDM patients lycopene concentrations were reduced. Expressed per before and during sitostanol ester-margarine treatment. Diabetologica 37, 773 ± 780. plasma lipid concentration, however, lycopene concentra- Gylling HK, Puska P, Vartiainen E & Miettinen TA (1996): Serum retinol, tion was not affected but (alpha beta)-carotene concentra- alpha-tocopherol, carotenes and lipid peroxide production during serum tion was decreased. In this present study similar results were cholesterol lowering by sitostanol ester margarine in a mildly hyperch- obtained. We observed a decrease in plasma (alpha beta)- olesterolemic population. Circulation 94, (Suppl) I-S578, abstract 3379. Haard PMM van, Engel R & Pietersma-de Bruyn ALJM (1986): Quantita- carotene, lycopene and alpha-tocopherol concentration after tion of trans-vitamin K1 in small serum samples by off-line multi- consumption of plant sterol enriched spreads. Correction for dimensional liquid chromatography. Clin. Chim. Acta 157, 221 ± 230. the reductions in the total plasma lipids, however, showed Heinemann T, Leiss O & Bergmann K von (1986): Effect of low-dose only plasma (alpha beta)-carotene concentration to be sitostanol on serum cholesterol in patients with hypercholesterolemia. reduced by about 10% after consumption of two of the Atherosclerosis 61, 219 ± 223. Katan MB, Zock PL & Mensink RP (1994): Effects of fats and fatty acids three spreads tested, that is the spreads enriched in the low on blood lipids in humans: an overview. Am. J. Clin. Nutr. 60, (Suppl), and the high dose of plant sterols. Carotenoids may have S1017 ± S1022. Cholesterol lowering effect of plant sterols HFJ Hendriks et al 327 Keys A, Anderson JT & Grande F (1965): Serum cholesterol response to Poppel G van (1993): Carotenoids and cancer Ð an update with changes in the diet. IV. Particular saturated fatty acids in the diet. emphasis on human intervention studies. Eur. J. Cancer 29A, 1335 ± Metabolism 14, 776 ± 787. 1344. Law MR, Wald MJ & Thompson SG (1994): By how much and how Tang JL, Armitage JM, Lancaster T, Silagy CA, Fowler GH & Neil quickly does a reduction in serum cholesterol concentrations lower risk HAW (1998): Systematic review of dietary intervention trials to of ischaemic heart disease? BMJ 308, 367 ± 373. lower blood total cholesterol in free-living subjects. BMJ 316, 1213 ± Ling WH & Jones PJH (1995): Minireview dietary phytosterols: a review 1220. of metabolism, bene®ts and side effects. Life Sci. 57, 195 ± 206. Vanhanen HT, Kajander J, Lehtovirta H & Miettinen TA (1994): LuÈtjohann D & Bergmann K von (1997): Phytosterolaemia: Diagnosis, Serum levels, absorption ef®ciency, faecal elimination and characterization and therapeutical approaches. Ann. Med. 29, 181 ± 184. synthesis of cholesterol during increasing doses of dietary Manson JE, Gaziano JM, Jonas MA & Hennekens CH (1993): Antioxidants sitostanol esters in hypercholesterolaemic subjects. Clin. Sci. 87, 61 ± and cardiovascular disease: a review. J. Am. Coll. Nutr. 4, 426 ± 432. 67. Mensink RP & Katan MB (1992): Effect of dietary fatty acids on serum Vanhanen HT, Blomqvist S, Ehnholm C, HyvoÈnen M, Jauhiainen M, lipids and lipoproteins. Arteriosclerosis and Thrombosis 12, 911 ± 919. Torstila I & Miettinen TA (1993): Serum cholesterol, cholesterol Miettinen TA & Vanhanen H (1994): Dietary sitostanol related to absorp- precursors, and plant sterols in hypercholesterolemic subjects with tion, synthesis and serum level of cholesterol in different apolipoprotein different apoE phenotypes during dietary sitostanol ester treatment. J. E phenotypes. Atherosclerosis 105, 217 ± 226. Lipid Res. 34, 1535 ± 1544. Miettinen TA, Puska P, Gylling H, VanHanen H & Vartiainen E (1995): Weisweiler P, Heinemann V & Schwandt P (1984): Serum lipoproteins Reduction of serum cholesterol with sitostanol-ester spread in a mildly and lecithin:cholesterol acyltransferase (LCAT) activity in hypercho- hypercholesterolemic population. N. Engl. J. Med. 333, 1308 ± 1312. lesterolemic subjects given beta-sitosterol. Int. J. Clin. Pharm. Ther. Pelletier X, Belbraouet S, Mirabel D, Mordret F, Perrin JL, Pages X & Tox. 22, 204 ± 206. Debry G (1995): A diet moderately enriched in phytosterols lowers Weststrate JA & Meijer GW (1998): Plant sterol-enriched margarines and plasma cholesterol concentrations in normocholesterolemic humans. reduction of plasma total-and LDL-cholesterol concentrations in nor- Ann. Nutr. Metab. 39, 291 ± 295. mocholesterolaemic and mildly hypercholesterolaemic subjects. Eur. J. Pollak OJ & Kritchevsky D (1981): Sitosterol. Monographs on Athero- Clin. Nutr. 52, 334 ± 344. sclerosis vol. 10. Basel: Karger. Weststrate JA & Hof HK van't (1995): Sucrose polyester and plasma Pollak OJ (1952): Prevention of hypercholesterolemia in the rabbit: carotenoid concentrations in healthy subjects. Am. J. Clin. Nutr. 62, successful prevention of cholesterol atherosclerosis. Reduction of cho- 591 ± 597. lesterol in man. Circulation 6, 459.